[<prev] [next>] [day] [month] [year] [list]
Message-Id: <20190920065429.19709-1-richard@nod.at>
Date: Fri, 20 Sep 2019 08:54:29 +0200
From: Richard Weinberger <richard@....at>
To: linux-mtd@...ts.infradead.org
Cc: linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
Richard Weinberger <richard@....at>,
Christoph Hellwig <hch@....de>,
Nicolai Stange <nicstange@...il.com>
Subject: [PATCH] ubifs: Remove obsolete TODO from dfs_file_write()
AFAICT this kind of problems are no longer possible since
debugfs gained file removal protection via
e9117a5a4bf6 ("debugfs: implement per-file removal protection").
Cc: Christoph Hellwig <hch@....de>
Cc: Nicolai Stange <nicstange@...il.com>
Signed-off-by: Richard Weinberger <richard@....at>
---
fs/ubifs/debug.c | 12 ------------
1 file changed, 12 deletions(-)
diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c
index a5f10d79e0dd..d67f91752f83 100644
--- a/fs/ubifs/debug.c
+++ b/fs/ubifs/debug.c
@@ -2737,18 +2737,6 @@ static ssize_t dfs_file_write(struct file *file, const char __user *u,
struct dentry *dent = file->f_path.dentry;
int val;
- /*
- * TODO: this is racy - the file-system might have already been
- * unmounted and we'd oops in this case. The plan is to fix it with
- * help of 'iterate_supers_type()' which we should have in v3.0: when
- * a debugfs opened, we rember FS's UUID in file->private_data. Then
- * whenever we access the FS via a debugfs file, we iterate all UBIFS
- * superblocks and fine the one with the same UUID, and take the
- * locking right.
- *
- * The other way to go suggested by Al Viro is to create a separate
- * 'ubifs-debug' file-system instead.
- */
if (file->f_path.dentry == d->dfs_dump_lprops) {
ubifs_dump_lprops(c);
return count;
--
2.16.4
Powered by blists - more mailing lists